PM and PdM work order compliance: whether your worst delays hit your most critical assets

A monthly lubrication route is scheduled every 30 days. In practice it keeps getting bumped, and the crew settles into running it nearer every five weeks. Nobody sounds an alarm, because the work does get done, just a little late. Then a quarterly vibration survey on a critical compressor slips three weeks, and nobody notices until the compressor trips one afternoon and takes the line down with it.

That is the gap this metric closes. It measures how far off schedule each preventive and predictive task is running, and then sorts that lateness by how critical the asset is, so you can see whether your worst delays are falling on the equipment that can least afford them.

What it actually measures

PM and PdM work order compliance is the percentage variance between the interval at which a task was actually completed and the interval it was supposed to run on. Each result drops into an overdue band, say within 10%, 10 to 25%, 25 to 50%, and over 50% late, and those bands are then cross-tabulated against each asset's criticality rank. The output is not just how late work is running, but whether the lateness is concentrated on high-consequence assets, which is the only version of the question worth answering. It is worth separating this clearly from plain on-time compliance, which simply asks whether a task beat its due date. This metric asks two harder things: by how much did the late ones miss, and did the misses land where they matter.

How to work it out

Overdue variance (%) = (Actual frequency / Planned frequency) × 100 - 100

A 60-day PM was last completed 78 days after the previous one.

Overdue variance = (78 / 60) × 100 - 100 = 30%

That 30% sits in the "25 to 50% late" band. On a criticality rank 1 asset it demands immediate investigation; on a rank 5 spare it only warrants a note. Repeat the arithmetic for every task in the period, tabulate by criticality, and the pattern, if there is one, jumps off the page.

The ten per cent rule

The widely used benchmark is the ten per cent rule: a task counts as properly compliant only if it is completed within ten per cent of its planned frequency, so a thirty-day task done between twenty-seven and thirty-three days is on time, and a quarterly one within about nine days either side. Aim for ninety per cent or more of tasks inside that window. At the very least, hold your critical assets to the rule even if you let lower-consequence work run a little looser, because that is where a slipped interval does real harm. Plants that enforce the ten per cent rule, especially on their critical equipment, tend to see measurably better reliability, for the simple reason that a disciplined, consistent interval is what makes a preventive task do its job in the first place.

The trap of a tidy variance

There is one failure mode this metric cannot see on its own, and it is worth naming. You can be perfectly compliant, every task on its interval and a tidy variance across the board, and still be losing. If a PM keeps an asset on a flawless schedule but the asset keeps failing anyway, the problem is not the timing, it is the task: the PM is aimed at the wrong failure mode, or the asset needs a different strategy altogether. A clean variance can quietly reassure you that all is well while you are, in truth, still reacting to breakdowns on a beautifully maintained machine. That is why this number has to be read next to the failure rate, never admired on its own.

Where it can mislead you

• Always cross the variance against criticality rank. The same delay means something completely different on a single-point-of-failure compressor than on a spare pump sitting idle.
• Measure it in calendar days for time-based tasks and in hours past interval for runtime-based ones; use the right basis for each asset or the variance is nonsense.
• A high overdue percentage on critical assets is an early warning of reliability risk. Escalate it before a failure proves the point for you.
• It tracks timing, not whether the task was done well, nor whether the task is even the right one. Pair it with an effectiveness measure and with the failure rate for the fuller picture.

A plain count of overdue tasks tells you that you are behind. This metric tells you something far more useful: whether being behind actually matters yet, by showing you whether the lateness is piling up on the assets you most need to protect, or merely on the ones that can comfortably wait.